The $11,000-Per-Employee Opportunity Your Competitors Don't See Yet

In 1995, the CEO of the startup Netscape, Jim Barksdale, stood in front of a room of British investment bankers at The Savoy in London. Netscape was about to IPO. The bankers wanted to know how the company would survive if Microsoft bundled a browser into Windows.

Barksdale gave them one line: "Gentlemen, there's only two ways I know of to make money: bundling and unbundling."

His CFO said the bankers thought he was crazy. But Barksdale had just described the fundamental cycle of business. Industries unbundle when specialists do one thing better than generalists. They rebundle when a new platform makes integration cheaper than coordination.

The bankers were right about the short term — Microsoft bundled the browser and killed Netscape. A decade later, the opposite happened.

Think about what SaaS actually did starting around 2005. Tomasz Tunguz at the venture capital firm Redpoint pointed out that nearly half a trillion dollars of market cap got built by unbundling Excel. Salesforce, Workday, Tableau, Carta, QuickBooks. Each one took a workflow someone managed in a spreadsheet and wrapped it in a better interface with fixed logic.

Every SaaS product is a spreadsheet that graduated.

Fifteen years later, we're about to go in reverse.

We have a limited number of spots left in the next OwnerRx AI for Business Owners Cohort starting April 13th. We’re offering $300 off for anyone who signs up before this Wednesday April 1st.

Claim your spot here →

Want to know if it’s worth it? Take it from your fellow business owner who participated in our last cohort:

“Your course has been… honestly changing our entire business and how we think about work. ”

Michael Raphael, CEO, Indevets.

AI rebundles most of it. And it changes who gets to do the bundling.

Enterprise software stocks have shed $285 billion in value since Claude Cowork launched. Software ETFs are down 30% this year. Publicis Sapient, a major consulting firm, just cut its Adobe licenses roughly in half and replaced them with AI tools. A services company eating a software company. That's not a one-off.

Andreessen Horowitz, one of the largest VC firms, published a piece in February arguing that vertical software will survive because domain-specific work requires compliance, multi-party workflows, and deep context. They're right about vertical. They might be wrong about who wins.

Here's what a16z missed. The vertical winners might not be software companies at all. They might be services firms that have the deepest client relationships and the best understanding of the customer's actual business.

The old world drew bright lines. You were either a services provider or a software provider. You picked a lane. A marketing agency didn't build its own analytics platform. An accounting firm didn't write its own audit software. You used Salesforce and HubSpot and Zendesk and QuickBooks and forty other tools, and your job was the judgment layer on top.

Those lines are dissolving. Claude Code is commoditizing software development. I've never written a line of code in my life and I've built several software products in the last six weeks. With new and more powerful models coming in the next month from OpenAI and Anthropic, it gets easier still.

The model companies themselves are already eating the basic productivity layer. Claude Cowork means you never have to open a Microsoft 365 or Google Workspace app again. Within 12–18 months, your software won't be viable unless it connects into Cowork.

That scared Microsoft enough to license Anthropic's product. They're releasing Copilot Cowork on May 1. This is the largest software company in the world shipping a flagship product built on a startup competitor's technology.

Some days I think Anthropic will eat everything. Claude will be the economy. That's obviously wrong. But here's the question worth asking: what's actually defensible?

Not general-purpose software. Not basic productivity. Not anything an AI agent can replicate in an afternoon.

What's defensible is specialized, vertically tuned work delivered by people who understand the customer's business and whom the customer trusts.

A16z thinks that means a SaaS startup. In many cases, it will. But there's another winner they're not talking about: you.

If you run a services firm with deep vertical focus, AI lets you replicate the software your clients use. Not theoretically. Right now. The customer support system. The project management tool. The reporting dashboard. The onboarding workflow. You can build agent-driven versions of all of it, tuned to your client's specific business, integrated with their actual data.

SMBs spend over $11,000 per employee per year on SaaS. They use an average of 253 apps. Nearly a third of that spend goes to seats nobody uses. That bloated vendor list isn't just inefficiency. It's your market.

And it's not just software spend. Every adjacent service your customers pay for is fair game too. Do I really need to pay a lawyer $600–1,200 an hour for 20 hours to develop a compliance plan? Or can I build an agent that does 95% of the work and pay a lawyer for a 30-minute review? You know the answer.

I'm narrowing in on a few candidate verticals for a play like this. I've started mapping every vendor that companies in those verticals pay — every software subscription, every outside services firm, every consultant.

I'm not starting with their ERP or their most regulated workflow. I'm starting where the fruit hangs lowest, adjacent to my current service. If I were doing marketing for CPA firms, I'd look at their sales process first. You're paying for Salesforce and using half the features? Let me replace it with a custom CRM tuned to exactly what CPAs need.

Stop thinking of yourself as a marketing firm. Start thinking of yourself as the company that enables CPA firms.

When one thing gets commoditized, the things around it go up in value. AI is commoditizing software. That means distribution and relationships appreciate.

Focus: pick your vertical niche. If you already have one, go deeper. Know your clients' businesses better than their other vendors do. That knowledge is your moat.

Bundle: start eating the spend. Every dollar your client sends to another vendor for something you could deliver with an AI agent is a dollar you're leaving on the table. Your ability to replicate what those vendors do is going to accelerate faster than most people expect over the next 12 months.

The companies with the deepest relationships and the most vertical focus win this. Not the ones with the best software.

The ones who know the customer best get to replace everything else.